The objective of the proposed study is to investigate possible mechanisms leading to differences in birth weight, duration of pregnancy and fetal growth between Black and White women. It focuses on variations in physical activity and social supports during pregnancy as a possible explanation for these racial differences, while simultaneously considering other variables known to vary between Black and White women. Physical activity during pregnancy has not been previously considered in studies of racial differences in pregnancy outcomes. The study sample is drawn from women attending five clinics associated with the University of Maryland Medical Systems (UMMS) for their prenatal care. They will be identified at their first prenatal visit between June, 1988 and November, 1989. All White women will be sampled and one in two Black women will be randomly selected for the study. Study participants will be interviewed between 19 and 26 weeks gestation and in the immediate postpartum period about their physical activity, social supports and smoking and alcohol use during pregnancy. The interview data are linked with data on demographic variables, medical obstetric history and complications of pregnancy available through the Maryland Perinatal Data Base (MPDB), a computerized data system including prenatal visit, labor and delivery, and postpartum data for the study women. Dependent variables include bith weight, gestational age and intrauterine growth retardation. Physical activity is measured in terms of activity at work, at home (including housework and child care) and exercise. Social supports focus on assistance in housework and child care as well as social and emotional supports. Other variables reported in previous research to vary between Black and White women that may impact on pregnancy outcomes are considered in the study. They include maternal demographic variables, health related behaviors, complications of pregnancy and medical-obstetric history. Multiple regression analysis will be utilized to study the effect of race and physical activity on the dependent variables. A stepwise analysis in phases will be performed to investigate the change in the coefficient for race and the goodness of fit of the model when physical activity, social supports and other known risk factors are added to the model.